A Watermarking Framework for Outsourced and Distributed Relational Databases

Rani, Sapana; Koshley, Dileep Kumar; Halder, Raju

doi:10.1007/978-3-319-48057-2_12

Cited by 5 publications

(1 citation statement)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The correct functioning of the technique requires the parameters having the same value in the embedding and the extraction processes; otherwise, the identification of the WM fails. Finally, the use of at least one parameter is required, being that the case of the secret key (SK), whose value should be known only by the data owner [13].…”

Section: Introductionmentioning

confidence: 99%

A Double Fragmentation Approach for Improving Virtual Primary Key-Based Watermark Synchronization

et al. 2020

View full text Add to dashboard Cite

Relational data watermarking techniques using virtual primary key schemes try to avoid compromising watermark detection due to the deletion or replacement of the relation's primary key. Nevertheless, these techniques face the limitations that bring high redundancy of the generated set of virtual primary keys, which often compromises the quality of the embedded watermark. As a solution to this problem, this paper proposes double fragmentation of the watermark by using the existing redundancy in the set of virtual primary keys. This way, we guarantee the right identification of the watermark despite the deletion of any of the attributes of the relation. The experiments carried out to validate our proposal show an increment between 81.04% and 99.05% of detected marks with respect to previous solutions found in the literature. Furthermore, we found out that our approach takes advantage of the redundancy present in the set of virtual primary keys. Concerning the computational complexity of the solution, we performed a set of scalability tests that show the linear behavior of our approach with respect to the processes runtime and the number of tuples involved, making it feasible to use no matter the amount of data to be protected.

show abstract

Section: Introductionmentioning

confidence: 99%